# Research

**April Heyward’s Research Sample RStudio Projects**

**Note: Move mouse to the bottom right hand of video for Fullscreen option** **after starting video.**

**ORCID iD**

**Research Interests**

- Bibliometrics
- Broadening Participation in STEM
- Data Science
- E-Government
- Machine Learning
- Public Administration
- Public Policy
- Research Methods
- Science and Technology
- Science Diplomacy
- Science of Science
- Science Policy
- Social Media Research

**Doctoral Research**

**Valdosta State University IRB 04125-2021 Protocol Exemption Approval**

**Quantitative Research – A Data Science and Machine Learning Approach to Comparative COVID-19 Policy Responses – (January 2021-Present)**

- Study aim is to employ Data Science and Machine Learning in the investigation of health and economic policy responses to the COVID-19 pandemic in developed countries and developing countries approved by Valdosta State University IRB.
- Prior research indicated more rigorous research is needed for contribution to the body of knowledge from a Public Policy perspective.
- Examining literature for the development of the research design and execution to include two-tailed (non-directional) research hypotheses, null hypothesis, independent variable(s), dependent variable(s), etc.
- Research aim is to answer how does Data Science and Machine Learning can inform Public Policy about the COVID-19 pandemic, what is the state of the COVID-19 pandemic in developed countries and developing countries, and empirically assess the correlation between policy responses and the state of COVID-19 in select countries.
- Employing secondary data from multiple data sources.
- Data wrangling, data analysis, and data visualization will be executed in RStudio.
- Initial data analysis will include time series analysis, text mining, sentiment analysis, and spatial data analysis.
- Creating R Scripts for reproducibility of data wrangling, data analysis, and data visualization in RStudio.

**Valdosta State University IRB 03979-2020 Protocol Exemption Approval**

**Quantitative Research – Measuring the Effectiveness of E-Government Delivery Models in Developed Countries and Developing Countries – (January 2020-Present)**

- Measuring the effectiveness of E-Government delivery models in developed countries and developing countries is the second phase of the quantitative study measuring the effectiveness of e-government delivery models approved by Valdosta State University IRB.
- Examined literature for the development of the research design and execution.
- Research aim is to answer what is the state of e-government delivery models in developed countries and developing countries.
- Employing secondary data from the first study phase plus developing two new datasets for data analysis.
- Integrating data science and machine learning algorithms for data analysis in R.
- Developed two-tailed research hypotheses (H
_{1}, H_{2}, H_{3}, H_{4, }H_{5}, H_{6}) and null hypothesis (H_{0}). - Creating R Scripts for reproducibility of data wrangling, data analysis, and data visualization in RStudio.
- Performing descriptive statistics with summary function and describe function.
- Creating boxplots to visually depict outliers with boxplot function and identified outliers with boxplot.stats function.
- Employing ANOVA with the aov function to identify significant differences and Tukey’s Honest Significant Differences with the TukeyHSD function to identify the location of significant differences.
- Employing K-means clustering with the kmeans function to cluster similar observations into groups and Hierarchal clustering with the hclust function to group observations based on similarities. Creating a dendrogram to depict the hierarchal relationships of the clusters as a tree diagram.

**Quantitative Research – Measuring the Effectiveness of E-Government Delivery Models from a Public Administration Perspective – (August 2019-Present)**

- Examined literature for the development of the research design and execution.
- Research aim is to answer what is the state of e-government delivery models globally.
- Employing a quantitative longitudinal design utilizing secondary data.
- Developed two-tailed research hypotheses (H
_{1}, H_{2}, H_{3}, H_{4, }H_{5}, H_{6}) and null hypothesis (H_{0}). - Imported datasets into RStudio with read_csv function.
- Loaded R packages: tidyverse, psych, stats, rmarkdown.
- Created tibbles with as_tibble function.
- Performed descriptive statistics with summary function and describe function.
- Performed Pearson correlation analysis with corr.test function, created correlation matrix with round(cor) function, and depicted correlation analysis on scatterplots with plot function.
- Created boxplots to visually depict outliers with boxplot function and identified outliers with boxplot.stats function.
- Created six simple linear regression models and two multiple linear regression models with lm function.
- Created regression diagnostic plots (residuals vs fitted, normal q-q, scale-location, residuals vs leverage) for each regression model with par function and plot function.
- Creating R Scripts for reproducibility of data wrangling, data analysis, and data visualization in RStudio.
- Preparing for scholarly publications and presentations.
- Citing R, RStudio, and R packages with citation function and RStudio.Version function for scholarly publications and presentations.

**Independent Research**

**Science Diplomacy: Examining the Relationship Between Science and Diplomacy**

**(May 2022-present)**

- Examining literature for the development of the research design and execution.
- Conducting Bibliometric Analysis, Machine Learning, and Natural Language Processing (NLP) algorithms in RStudio employing .bib and .txt data from the Web of Science Core Collection.
- Prior research reveals heavy focus on the diplomacy side of the science diplomacy equation. Science must be equally understood including how science works which also invites the discussion of the science of science. This can extend comprehension of science diplomacy.
- Research aims are to underscore the role of science in science diplomacy, examine the relationship between science and diplomacy, and to demonstrate the broader impacts of science diplomacy.
- Investigating the three primary aspects of science diplomacy: diplomacy for science, science for diplomacy, and science in diplomacy.

**Science Policy: Examining the Relationship Between Science and Policy** –** (February 2022-present)**

- Examining literature for the development of the research design and execution.
- Conducting Bibliometric Analysis, Machine Learning, and Natural Language Processing (NLP) algorithms in RStudio employing .bib and .txt data from the Web of Science Core Collection.
- Investigating the relationship between science and policy which also invites the discussion of the science of science.
- Research aim is to investigate how does science inform public policy.

**Quantitative Research – A Public Administration Research Approach and Empirical Perspective of COVID-19 – (March 2020-present)**

- Examining literature for the development of the research design and execution.
- Research aim is to employ data science and machine learning algorithms in a public administration research approach and empirical perspective of COVID-19.
- Time series analysis will be employed in R to examine the economic impact of COVID-19 related policies.
- Text mining will be employed in R of COVID-19 research literature for text processing, modeling, analysis, and visualization.
- Naïve Bayes will be employed in R for text classification.
- Sentiment analysis will be employed in R to examine public opinion of COVID-19 related policies.
- Spatial analysis will be employed in R of COVID-19 longitude and latitude cases.

**Quantitative Research – Integration, Challenges, and Future Direction of Data Science in Public Administration – (August 2019-present)**

- Examining literature for the development of the research design and execution.
- Research aim is to answer what is the state of data science in public administration.
- Integrating data science and machine learning algorithms for data analysis in R.
- Loaded R packages: tidyverse, psych, stats, rmarkdown.
- Initial data collected and wrangling dataset A with 23,859 observations and dataset B with 16,716 observations in R.
- Importing datasets with read_csv function in RStudio.
- Creating tibbles with as_tibble function.
- Skipping rows with skip = function.
- Deleting variables with select function.
- Combining columns with unite function.